coons_c_weber_m_eds_paternalism_theory_and_practice
.pdfIntroduction: Paternalism – Issues and trends
Christian Coons and Michael Weber
John Stuart Mill famously decried paternalism of any kind, whether it is carried out by private individuals or the state:
[T]he sole end for which mankind are warranted, individually or collectively, in interfering with the liberty of action of any one of their number, is self-protection
. . . the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or moral, is not a su cient warrant.1
Equally famously, critics have charged that Mill’s utilitarianism is inadequate to ground such an absolute prohibition. If there is a case for an absolute ban on paternalism, many have thus thought, it must be grounded instead in the fundamental importance of one of a family of considerations that includes liberty, freedom, and autonomy. However, it is unclear whether this move really helps, or so we will suggest.
This mere snippet from the historical debate is enough to show that paternalism is a topic that engages deep philosophical issues in normative ethics and political philosophy, including the significance and nature of freedom and autonomy, and the relation between individuals and the state. But interest in the topic is of course not primarily due to its theoretical depth. Instead, it is due to practical considerations. A wide variety of policies and laws in the United States and elsewhere are considered paternalistic, and are controversial for just that reason. For instance, there are laws that require motorcyclists to wear helmets and passengers in cars to wear seatbelts. Government agencies regulate both prescription and recreational drugs. Taxes are levied on cigarettes, and bans on trans-fats have been enacted. Participation in pension programs, such as Social Security in the United States, is mandatory. Mill’s negative view of paternalism is reflected in some identifying policies along these lines as part of an ever-growing “nanny-state.”
1 Mill, On Liberty, 9.
1
2 christian coons and michael weber
Because paternalism in this way raises significant theoretical and practical concerns, it has been a topic of long-standing interest to moral and political philosophers as well as political actors and the lay public.
Interest has been heightened recently due to an alleged breakthrough, a way of squaring paternalism with libertarian concerns for that family of values that includes liberty, freedom, and autonomy. This “libertarian paternalism,” championed by Richard Thaler and Cass Sunstein in their recent bestseller, Nudge: Improving Decisions About Health, Wealth, and Happiness, raises new and interesting questions, especially in virtue of its appeal to recent research in psychology and behavioral economics.2 The essays in this volume address the fundamental and long-standing issues raised by paternalism as well as the debate sparked by Thaler and Sunstein. In this introduction we will not summarize each of the contributions to the volume. Instead, more helpfully we hope, we will touch on some of the key aspects of the present-day discussion of paternalism, so as to provide a broad context for thinking about the essays herein. Our discussion of these new developments will be framed by a central question in both classic and contemporary debates: What, if anything, makes paternalism morally problematic? Where appropriate, of course, we will indicate how specific contributions to this volume figure in the larger narrative.
Normative debates about paternalism – or at least “hard” paternalism – don’t usually concern whether it is problematic but rather how problematic it is.3 Specifically, most assume that we have a pro-tanto reason to avoid paternalism towards competent adults. The real debate concerns if it can nevertheless ever be justified, and if so, when. Despite this, in recent years it has become surprisingly di cult to explain exactly why paternalism is even pro-tanto objectionable, and more di cult still to defend its absolute prohibition. It has become more di cult not because the world has somehow changed but rather because philosophical reflection has made it clearer what paternalism does and does not involve. Crucially, there’s a growing consensus that it needn’t be regarded as coercion, removal of
2Thaler and Sunstein, Nudge.
3The soft paternalist maintains that interference is justified if and only if the “target” person is not acting su ciently knowledgeably and voluntarily. Although soft paternalism is sometimes understood as paternalism that does not impose values or ends on its targets, that is not how we’re using it here. Hard paternalists allow interference to benefit agents acting knowledgeably and voluntarily. Our discussion is concerned with hard paternalism. As such, from here forward “paternalism” and its derivatives should be taken to refer to hard paternalism unless specifically noted; indeed Feinberg, Harm to Self, 16, famously questions whether soft paternalism qualifies as paternalism at all.
Introduction: Paternalism – Issues and trends |
3 |
choice, or even disregard for the target’s evaluative perspective. Because of this, some traditional views about why paternalism is pro-tanto wrong are no longer available. We make this point not to advocate paternalism, but to help illustrate why despite all the attention previously paid to it, it remains a rich and evolving topic for discussion.
Of course, one might skeptically insist that it is paternalism’s very richness that makes it problematic as a focus of direct study. After all, in paternalism, perhaps more than any other topic in moral and political philosophy, deep conflicts between competing traditions emerge, not only about the limits of state authority, but in more general disputes about the relative priority of well-being, freedom, choice, and autonomy, and whether these are ideals to be promoted or respected. The skeptic about the direct study of paternalism might then insist that no progress can be made without first resolving these deeper disputes. But this assumes what we might call a “top-down” approach: General philosophical issues must be solved first, the results of which are then simply applied to ground-level disputes in political philosophy or applied ethics. The top-down approach, however, is both methodologically questionable and di cult to consistently maintain. For example, it would presumably be a reductio of any broad theoretical view if it entailed that we may not stop a friend from drinking the gasoline she thinks is gin. So our verdicts about paternalism needn’t be a mere application of our more general commitments in moral and political philosophy. Rather, the opposite may be true: Careful reflection on paternalism’s moral status may help illuminate or adjudicate debates about those deeper issues. Indeed, if we see the coercive power of the state as justified in part by its claim to benefit those subject to its power, it appears state authority itself rests on a kind of paternalistic rationale. Paternalism’s relevance to wider debates is also apparent in Christopher Wellman’s complaint that Rawls’ Principle of Fairness is objectionably paternalistic; Richard Arneson’s contribution in this volume defends Rawls on this score.4 In this debate, paternalism’s status is treated as determining, rather than being determined by, our more general normative commitments.
So just as consensus at the theoretical level may be useful in drawing conclusions about how we should regard paternalism, consensus about paternalism may inform our more general theoretical positions.5 With that
4Wellman, “Samaritanism and the Duty to Obey the Law” in Wellman and Simmons, eds., Is There a Duty to Obey the Law?
5A powerful instance of this sort of strategy can be found in Wall, “Self-Ownership and Paternalism.”
4 christian coons and michael weber
in mind, it is worth returning to the widely held view that paternalism is inherently problematic – that there is always at least some pro-tanto reason to avoid it. But what, precisely, makes paternalism problematic? And what does this reveal about deeper moral views? At least this much initially seems obvious: The problem does not lie in its end alone – in the fact that paternalism is undertaken for the good of or the sake of its targets.6 Paternalism appears to be very broadly speaking benevolent. But there’s nothing wrong with benevolence per se. What makes paternalism at least pro-tanto wrong, then, presumably has to do with the means – with how it goes about benefiting its target. For example, forcing you to have a second helping of broccoli casserole – a tasty and healthy choice – is objectionable to the extent that it is because I am forcing you to do it. It would likewise be objectionable, and apparently for the same reason, if I forced you to eat it for a scientific experiment I am conducting. What is pro-tanto wrong with paternalism in this case, apparently then, has nothing to do with the fact that I am trying to improve your health or please your palate. That it is good for you may speak in its favor. That it promotes my scientific experiment seems to count less, most likely because I am the beneficiary rather than you who are forced to eat the casserole. In both cases, however, it seems most plausible that the reasons to force you to eat the casserole are too weak to outweigh the reason not to coerce you. Generalizing from this example, we might suspect that what’s wrong with paternalism is rather straightforward: Paternalism involves some form of coercion or interference, which requires special justification. Advocates of a general prohibition on paternalism are, then, simply those who think this justifi- cation cannot be met – the well-being of the target of paternalism never outweighs the morally problematic use of coercion.
But such an account will not do. As is now more widely recognized, omissions can be paternalistic, and thereby problematic, even though the omission would not otherwise require any justification. Here, fascinatingly, paternalism’s benevolence appears to contribute to what makes it problematic. To illustrate, we may omit telling you about tonight’s concert because we believe that we ought not tempt you. This omission may require no justification: if we had simply not felt like telling you, we would do nothing objectionable. But when we omit for your sake, it seems
6For simplicity, we will generally speak of paternalism and paternalistic acts, though it is probably more accurate to describe act–reason pairs as paternalistic, as Grill, “The Normative Core of Paternalism,” emphasizes. As will become clearer below, we believe there’s an important distinction between acts performed for a target’s well-being and those performed for her sake.
Introduction: Paternalism – Issues and trends |
5 |
you have greater grounds to object. Specifically, our omission seems to involve reasoning for you – we’ve weighed the options for you, and decided not to tell you partly because we believe you might reach a di erent and “imprudent” conclusion. This feature of paternalism – a sort of reasoning for another – has not always been explicitly highlighted, but takes center stage in Seana Shi rin’s recent but already highly influential characterization of paternalism. On Shi rin’s view, paternalism by A towards B is behavior (including omissions) meeting the following conditions:
(a)aimed to have (or to avoid) an e ect on B or her sphere of legitimate agency
(b)involving the substitution of A’s judgment or agency for B’s
(c)directed at B’s own interests or matters that legitimately lie within B’s control
(d)undertaken such that, compared to B’s judgment or agency with respect to those interests or other matters, A regards her judgment or agency to be (or as likely to be), in some respect, superior to B’s.7
This view in fact highlights two features of paternalism that have historically been overlooked or under-emphasized: i) paternalism may not involve coercion or active interference; ii) it needn’t involve a specific concern for the target’s well-being. The definition is not merely a potential theoretical advancement, it also sheds light on types of paternalistic practices and policies that, historically, have gone unappreciated. Daniel Haybron and Anna Alexandrova’s contribution to this volume is an excellent illustration of the point, as it uses some of Shi rin’s insights to turn the tables on those who argue happiness-driven economics is objectionably paternalistic in a way that more traditional “minimalist” methodology in economics is not. And Sigal Ben-Porath’s contribution uses some of these same insights to shed light on the paternalism that is at the heart of contemporary debates about school choice.
Shi rin’s account also o ers us a direct answer regarding paternalism’s morally problematic nature. There’s no need to explain why paternalism is presumptively problematic, as all one needs to do is possess the concept to see that it is. For on her view “paternalism” is a kind of thick negative ethical term: Paternalistic acts by definition intrude on areas the target legitimately controls. Put in laymen’s terms, the paternalist intrudes in areas that are “none of her business” – it is an invasion of personal sovereignty. Return to our omitting to tell you about the concert. When
7 Shi rin, “Paternalism, Unconscionability Doctrine, and Accommodation.”
6 christian coons and michael weber
we omit because we don’t feel like telling you, that’s “our business,” but given that we omit out of exclusive concern for you, we’ve e ectively made your business – questions about what is best for you – our business.
Though there’s much to learn from Shi rin’s account, it also seems dubious that paternalism’s problematic nature is part of its conceptual content. In the first place, some, like Thaler and Sunstein, think that paternalism is sometimes not problematic at all – that the salient negative features are not just outweighed, but eliminated. And as Peter de Marne e emphasizes in his contribution to this volume, paternalism need not violate an individual’s “sovereignty” – matters over which one should have legitimate control. Of course, we might contest these conclusions. But Shi rin’s view entails that they are conceptually incoherent, which doesn’t seem right. More strikingly, notice that if Shi rin were right, error theorists and moral skeptics should deny paternalism even exists. And although there have been some quibbles about which act-types are genuinely paternalistic, we tend to agree about which particular acts and policies qualify. This agreement would seem miraculous on Shi rin’s view, as there’s great variation in our views about which matters are legitimately ours to control, and so there should be equally great variation in our views about which acts are paternalistic. Indeed, a hardline, old-fashioned act-consequentialist can regard some measures as paternalistic, and have reservations about them as such. But such theorists may also deny that there are any matters that legitimately lie within an agent’s control because legitimate control implicitly appeals to something akin to an entitlement or right – normative categories that such a consequentialist may not countenance. Presumably, we should not characterize paternalism in a way that entails these consequentialists must deny its existence.
We don’t intend to provide our own definition of paternalism here. Rather we wish only to raise a concern about normative or morally loaded characterizations.8 If we are to explain what is pro-tanto wrong with paternalism we may be better o identifying a non-normative characteristic of such acts that is morally problematic. Shi rin’s “substitution of judgment” in some non-normatively specified domain(s) – perhaps the very ones Shi rin believes are “within our legitimate control” – is an intriguing and promising possibility, not merely to better understand paternalism but to cast light on what is now emerging as a central but neglected morally relevant feature – a feature whose moral status will be
8 For a very nice survey of recent definitions of paternalism including his own normative characterization, see Grill, “Anti-Paternalism and Public Health Policy.”
Introduction: Paternalism – Issues and trends |
7 |
touched on again below. Nevertheless, we should be further wary of canonizing Shi rin’s influential characterization. For Daniel Haybron and Anna Alexandrova’s contribution o ers some compelling reasons why it needs to be at least modified, and deeper, more developed concerns are a focus of Gerald Dworkin’s opening section.
If we set aside morally “loaded” characterizations of paternalism, we again face the di culty of explaining why paternalism is at least pro-tanto objectionable, and why some have thought it never permissible. Ironically, paternalism’s most famous critic, Mill, appears to be of little help. Mill is traditionally seen as an exponent of a tradition – act-utilitarianism – which has a particularly di cult time explaining why paternalistic acts are presumptively problematic.9 The trouble, of course, is that whatever the correct characterization of paternalism, there can’t be a guarantee that such acts never maximize utility. Paternalistic acts also need not involve any (perhaps outweighed) loss of utility. Accordingly, this tradition cannot explain either why paternalism is always wrong or even why there’s always some reason to avoid it.
Utilitarian anti-paternalists might insist that this criticism involves a too-narrow view of what it is to make a presumptive case against paternalism. Rather than trying to locate a problem with what paternalists do, they might say, we should look at the relation between the paternalist and what she does. More specifically, one might argue that given our epistemic position, paternalism can always be expected to yield suboptimal outcomes. Following Mill, perhaps the proposed target of paternalism is always better placed to know where her happiness lies and how to best achieve it. And although we might quibble with this presumption – e.g., if I’m depressed, a wise friend may be a better judge than I about where my self-interest lies – Mill’s claim is quite plausible, and it is especially powerful against state paternalism. State executives, assemblies, and state bureaucrats do not know us well – are not wise friends – and so are especially ill-placed to know (better than you) what’s good for you. Furthermore, because state paternalism manifests in policies and statutes that apply to the citizenry at large, they are too blunt to be sensitive to the di ering interests of the citizens bound by the policy or statute. Paternalist penal statutes su er from still further justificatory problems, as Douglas
9Mill’s actual view is almost certainly more rich and nuanced than the traditional hedonistic act– consequentialism he’s often portrayed as holding. See, among many others, Dworkin, Mill’s On Liberty; Arneson, “Mill Versus Paternalism”; and more recently Jacobson, “Utilitarianism Without Consequentialism.”
8 |
christian coons and michael weber |
Husak explains in his contribution to this volume. The utilitarian antipaternalist, then, could argue that the problem is that paternalism rarely promotes the good, and the paternalist is never in a position to know when it will. Indeed, these facts would be su cient to make paternalism always wrong on a rule-utilitarian view and versions that appeal to expected rather than actual consequences.
The worry about such a reply is that even if the host of sociological, psychological, and economic data required to make such a case were available, it would seem useful only as further rationalization and not a reflection of what’s seemingly wrong with paternalism. After all, we tend to be anti-paternalists independent of any grasp of the empirical data required to make this utilitarian case for it. Notice that the case focuses on an appeal to paternalism’s ine cacy or “backfiring.” But this focus seems misplaced, because anti-paternalists do not generally lament the ine ectiveness of these policies; they wouldn’t wish or seek a “breakthrough” in sociopsychological engineering that would allow us to craft successful paternalistic policies. In fact, as Jeremy Blumenthal’s essay in this volume points out, new empirical work is revealing that paternalism is e ective in some contexts. But even so, its e ectiveness is presumably rightly regarded as merely a necessary and not su cient condition for its justification. Apparently then, conventional anti-paternalism, and the nearly uncontroversial sentiment that paternalism is in need of justification, reflects some nonutilitarian elements of our moral thought.
Utilitarianism’s inability to capture what is seemingly wrong with paternalism might not be a feature of consequentialism generally. Perhaps a consequentialist will fare better by looking beyond happiness and towards other values such as freedom and autonomy. However, it is unclear this helps. For even if the target of a paternalistic act necessarily thereby loses some freedom or autonomy, such limitations can ultimately produce a net increase in freedom or autonomy. Mill’s lone exception to his anti-paternalism illustrates the point well: Preventing one from contracting into slavery increases one’s freedom overall.10 So merely appealing to freedom and autonomy as values to be promoted cannot capture what is wrong with paternalism, but instead will yield exceptions. In this volume, Larry Alexander discusses voluntary slavery contracts in great detail, ultimately o ering, contra Mill, a provocative defense of such contracts.
10 Arneson, “Mill Versus Paternalism,” has a particularly interesting discussion of Mill’s case of contracting into slavery.
Introduction: Paternalism – Issues and trends |
9 |
In general, it seems that no matter what value the consequentialist takes on board, the problem will be the same. Consider, for instance, certain values that seem perfectly suited for anti-paternalism, values such as noninterference and non-paternalism itself. Each of these could be promoted by paternalism. For instance, the State Department might place travel restrictions to various countries on citizens by appeal to the value of noninterference: Interfering with your travel will save you from even greater interference abroad. Similarly with non-paternalism: Acting paternalistically towards you now could save you from much greater paternalism later. For instance, if you live in the world’s least paternalistic state, a prohibition on emigration may maximally promote non-paternalism.
It seems, then, that the tension between consequentialism and deep anti-paternalism is rooted in consequentialism’s view of all values, including liberty, freedom, and autonomy, as things to be promoted, where promotion involves maximizing. Deontologists, in contrast, sometimes suggest that such values are to be respected, or that people are to be respected as free and autonomous. But what does it mean to respect someone as free and autonomous? What does such respect require? And most importantly, how might paternalism fail to fully respect persons as free and autonomous? Perhaps if we want to understand ordinary antipaternalist sentiment, it make sense to turn to ordinary expressions of it. Typically, we say things like “you/the government don’t know what’s good for me,” “you/the government can’t tell me what’s good for me,” or “it’s none of your/the government’s business.” Some of these expressions won’t help, as they are simply ways of expressing anti-paternalism or reasons why it may be ine ective. However, some may be fruitful. So, for instance, the expression “you/the government can’t tell me what’s good for me,” suggests that paternalism fails to respect persons as free and autonomous by imposing values on them. For instance, consider laws requiring motorcyclists to wear helmets. Such laws apparently assume that the risk of injury or death outweighs the thrill of riding without a helmet. But some motorcycle enthusiasts might well value the thrill over added safety. Consider also travel bans to dangerous countries such as Somalia, or Syria during the 2012 uprising. It may be true that such travel isn’t what is best for you, even by your own lights. However, the ban assumes that you attach greater significance to your own well-being than you do to, for instance, documenting the events in such war-torn countries. The ban, then, imposes at least a weighting, attaching greater significance to personal well-being than other ends. But surely some, e.g., brave war-correspondents, do not weigh their personal well-being quite
10 christian coons and michael weber
so highly. The problematic feature of paternalism, then, might be “replacing” the values of the target with the values of the paternalist.
But paternalism need not involve the imposition of values. This becomes apparent if we remind ourselves that people often fail to do what best reflects their values. Consider two ways this might happen. First, a person might choose inadequate (or suboptimal) means to their ends, e.g., optimistically leave only twenty minutes to travel from home to the airport when in all likelihood it will take longer. Someone – perhaps the prospective traveler’s spouse – might coercively interfere in such cases so as to ensure that the person (e ciently) realizes his ends, e.g., by removing his breakfast plate before he is done. This seems paternalistic, as it is done for the traveler’s sake. However, it needn’t involve imposing values.
More strikingly, our potential traveler might be akratic. It is of paramount importance to him to get to the airport to make his flight. He knows that a leisurely breakfast would be nice but is not as important as making his flight. And yet he finds himself lounging over breakfast despite the whispers of his conscience. In this case, too, someone might coercively intervene for his sake. And here it is only the paternalizer who seems to duly respect the potential traveler’s values. Though it is less explicitly discussed, this may in fact be paternalism’s most common form. It is very much like the Kantian paternalism in Michael Cholbi’s contribution to this volume, or the “loose” paternalism described by Dan Scoccia in his. The paternalist, of course, is meant to see himself as acting for the target’s sake, and it’s di cult to sincerely believe you’re doing so while knowingly imposing alien values on the target. If that’s correct, it’s not merely that paternalism does not necessarily “impose” values – the paternalist is also, in some sense at least, committed to avoiding doing so. If paternalism is even pro-tanto wrong, it does not appear to be because it involves failure to respect its target’s values.
Alternatively, one might think that respecting someone as free and autonomous has nothing to do with anything as rarefied as “respecting her values.” Paternalism is not suspect for this reason. Instead, what is suspect about paternalism is more on the ground-level: Provided I’m competent and acting voluntarily, any choice or decision I make, assuming it doesn’t threaten others or violate others’ rights, is mine to make. What requires respect is not my “values” but my decisions – whether or not they reflect my values.
This position certainly has a resonance and more than its share of popular currency. But it has at least three odd features. First, it’s hard to see how a practically rational person can really insist on such a thing. A practically rational person must assess their choices in terms of their values. Consider the first-person point of view. From this point of view we